Many stroke survivors have limitations in motor ability and compromised quality of life. Therapeutic interventions designed to enhance motor function and promote independent use of an impaired upper extremity following stroke are limited. There is a need to translate unique behavioral and cognitive techniques shown to have an impact on plasticity in the nervous system into practical, evidence-based, therapeutic interventions. One technique, constraint induced therapy (CIT), has produced results that can substantially reduce the incapacitating deficits of many patients with stroke and can increase their independence. Recent studies have shown that a complementary treatment, mental practice, can improve the performance of motor skills. Investigators have proposed the use of mental practice in physical rehabilitation as a cost-efficient means to promote motor recovery after damage to the central nervous system. The primary purpose of the proposed study is to collect data in an effort to estimate the clinical effectiveness of implementing one form of CIT, repetitive task practice (RTP) in conjunction with mental imagery training (MIT) to improve upper extremity motor function and quality of life of chronic stroke patients. Therefore, we hypothesize that mental imagery of a motor task combined with RTP may lead to decreased upper extremity (UE) impairment and improved UE functional outcome compared to RTP alone, and that common neural structures can be investigated using fMRI. Specific aims will seek to estimate: 1) the effectiveness of using RTP in conjunction with MIT on motor recovery of patients with stroke, 2) the effectiveness of using RTP in conjunction with MIT on health related quality of life of patients with stroke and 3) the association between changes in the organization of cortical activation maps during the performance of a specific task (executed vs. imagined finger flexion-extension sequence) and improvement in clinical motor function of patients with stroke when receiving RTP + MIT compared to MIT alone.